July 31, 1913] 



NATURE 



56: 



EXPERIMENTAL CANCER RESEARCH. 



EXPERIMENTAL cancer research can be 

 undertaken with two main objects in view : 

 (i) The investigation of the origin of cancer; 

 (2) The study of the properties and life-history of 

 the developed tumour. In regard to the first of 

 these objects the staff of the Imperial Cancer 

 Research Fund has confined its attention to the 

 question of heredity. In the present report : there 

 is a summary of observations on this point which 

 confirm the conclusions previously published. In 

 mice with recent cancerous ancestry, 20 per cent, 

 of all deaths were due to cancer, while in those 

 with remote cancerous ancestry the ratio was 

 n"6 per cent. It will be interesting to see 

 whether further research in this direction will 

 reveal any relation between the types of tumour 

 in the ancestors and their progeny, and whether 

 there will be any appearance of the Mendelian 

 phenomenon. 



The main efforts of experimental cancer research 

 are, however, directed towards an investigation 

 of the properties and life-histories of spontaneous 

 and propagated cancer. In this direction the 

 record is one of steady progress ; the scope of 

 investigation is continually expanding, and new 

 problems are constantly offering themselves for 

 solution. 



Perhaps the most striking phenomenon in ex- 

 perimental cancer research is the appearance of 

 a sarcoma during the propagation of a carcinoma. 

 Recent work brings out the interesting fact that 

 the power of inducing sarcoma formation is not 

 necessarily a permanent property of a particular car- 

 cinoma, but may be only transitory. Two tumours 

 have been previously observed to have this property. 

 In one of these, sarcoma formation was irregular, 

 and was promoted by rapid repetition of trans- 

 plantation. In the other, sarcoma formation 

 occurred with remarkable regularity if the tumours 

 were allowed to grow for about two months be- 

 fore transplantation, but did not occur if trans- 

 plantation was effected rapidly. The recent history 

 of these tumours has shown that the first has 

 completely lost the power of inducing sarcoma 

 formation. Of the second tumour four carcino- 

 matous substrains have been kept growing. In 

 one of these, sarcoma development has occurred 

 earlier at each transplantation until it has 

 occurred so early that it has been impossible to 

 retain the carcinoma in propagation, and this 

 strain has become purely sarcomatous. In another 

 strain the appearance of sarcoma has become pro- 

 gressively later until the power of inducing sar- 

 coma formation has been lost. In the other two 

 strains the same process is apparently going on. 



In this phenomenon of sarcoma formation there 

 thus await consideration not only the problem, 

 What causes the production of a sarcoma during the 

 propagation of a carcinoma? but also the prob- 

 lems, Why does a tumour which at one stage of 

 its propagation shows the power of inducing 



1 Eleventh Annu.il Report of the Imperial Cancer Research Fund. Pre- 

 sented July 24. 



NO. 2283, VOL. 91] 



sarcoma formation subsequently lose this 

 property? and What is the explanation of the 

 variability in this power displayed by different 

 strains of the same tumour? All we can say at 

 present in answer to these problems is that the 

 sarcoma formation is not due to any filterable 

 virus or other material derived from the carcinoma 

 cells, but requires the presence of living cells for 

 its induction. 



The curious variability of the properties of 

 mouse-cancer is also shown in other directions. 

 In some tumours the structure remains constant 

 during propagation for several years, while others 

 show great variability. An acinous carcinoma 

 may become alveolar in type ; a tumour originally 

 showing keratinisation, sebaceous transformation, 

 or glycogen formation may lose these properties 

 temporarily or permanently. Similarly in its 

 power of growth a tumour may show constancy 

 or variability. From a single tumour two strains 

 may be isolated, one of which constantly dis- 

 appears after a short period of growth, while the 

 other regularly grows progressively and gives 

 rise to metastases. The power of growth in a 

 particular tumour depends inversely on its power 

 of inducing in the host resistance to its own 

 growth. 



In all these cases of variability the question 

 arises, Does the variability reside in the tumour- 

 cells, or in the tissues of the host? The available 

 evidence goes to show that the variability centres 

 in the tumour itself. When, for instance, a slow- 

 growing tumour arises from one which grows 

 rapidly, when the structure of a tumour changes 

 in the course of propagation, or when a sarcoma 

 arises during the propagation of a carcinoma, 

 the changes must be ascribed to variations in the 

 tumour-cells, Ihe transplantation subsequently 

 effecting the isolation of the different characters 

 evolved. It is well to note that this variation 

 is not always in one direction. The change in 

 the tumour-cell is not always from one of lower 

 to one of higher differentiation and vice versd. 

 Also tumours showing high differentiation are 

 not necessarily the most constant, nor those of 

 low differentiation the most variable. 



One important point should be noticed ; that is, 

 the increasing evidence as to the identity in nature 

 of the mouse-cancer with the human disease. 

 Every feature of cancer in mice finds its parallel 

 in man. Squamous-celled carcinomata in mice 

 sometimes show the formation in them of cysts 

 lined with typical squamous epithelium, and this 

 property may remain constant during propagation. 

 The same feature is frequently found in human 

 squamous-celled carcinomata, and is repeated in 

 the metastatic growth. Similarly sarcoma forma- 

 tion may take place in human carcinoma, as it 

 does in propagated and spontaneous carcinoma 

 in mice. Again, in man the structure of the 

 primary tumour is often repeated with great con- 

 stancy in the metastatic tumours, whereas in 

 some cases there may be great variability, and 

 tumours of the same structure in man may show 

 threat differences in the power of growth, one 



